Description

Overview

Trading at merely ~5x EBITDA, small-cap electronic components manufacturer Bel Fuse (BELFB) has 150%+ upside over the next 3-5 years through a one-two combo of 1) their margins doubling from LTM levels while still having further room to improve to peer levels and 2) their EBITDA multiple increasing 50%+ as a result of item 1. 

The stock has already almost doubled year-to-date in a very challenging macro environment, due to strong organic improvement in results, but we think it remains substantially undervalued.  

We think Bel offers a little something for everyone and are thrilled to post it as our first idea on VIC - to start with, it’s inexpensive at 6x / 5x LTM / forward EBITDA.  But Bel also offers a compelling combination of long-term secular tailwinds (“electrification” of everything from grills to tractors) and a macro-independent, credible self-help story with lots of “low-hanging fruit” that should allow its EBITDA margins to expand substantially over the next 3 years even in the face of potential cyclical headwinds.  

The opportunity exists because the company spent more than a decade largely ignoring profitability and organic growth, instead primarily growing via acquisition (poorly).  While pockets of the business were highly profitable, other pockets of the business - almost unbelievably - were operating at negative gross margins due to a complete lack of visibility into SKU or customer-level profitability, and a culture and sales compensation process that were not focused on profitability.  

Under a new CFO, Farouq Tuweiq, who is young and spearheading change, the company has refocused on profitability rather than revenues, and has made significant strides already in improving margins and delivering organic growth.  While these initial successes have caused the stock to rerate quite a bit, there’s still a lot farther to go.

In fact, we think this is the rare story that is actually more attractive to us at a higher price than it was at a lower price, as the last few quarters have “proved out” the thesis.  There seem to be fewer risk factors today than we first started looking at it earlier this year - we started buying it in the teens and have continued buying it up to the current prices of around $25 - $26 per share, as the risks of failure have, in our view, substantially receded.  

We think that this company-specific story is particularly compelling in light of the broader economic concerns of today; we believe that Bel is well positioned to deliver positive absolute returns over the next 12-24 months even if the broader economy and market are challenged.

Brief Company Background / Why The Opportunity Exists

Before we get to discussion of the company’s products, business lines, and so on, we think it makes sense to start with some more qualitative elements of where the company has come from and where it is going.

We have a “type” when it comes to investing.  We typically look for quality small-cap businesses with a history of strong cash flow, at least modest growth, thoughtful capital allocation, and so on.

Small-cap electronic components company Bel Fuse (BELFB), with ~$600 million in trailing revenue, has long been the opposite of that - in fact, anyone who takes a cursory glance at the long-term history of the business would see absolutely no reason to invest in it.  The company’s median and average ROIC from 2012 - 2022 was a pitiful 2.9%.  Trailing revenue and EBITDA were lower at the start of 2022 than they were in 2015 despite spending $50 million on acquisitions in the interim.  

It looks even worse if you go farther back; EBITDA in 2021 was only 20% higher than 2005-2006 levels even though the company did roughly $300 million of M&A over that timeframe, nearly tripling revenue in the process (the focus on revenue over profits is a consistent theme here).  

Despite a massive bull market, if you had bought a share of BELFB at the start of 2010, you would have spent more time in negative territory than in positive territory through the ensuing 12 years.  Bel Fuse has been the opposite of a compounder and we cannot identify a single historical metric over the past 10-15 years that would make an investor think this company was worth investing in.

This depressing history is probably why Bel has never had a single writeup on Value Investors Club, and why we think - even after nearly doubling year-to-date - it is an underfollowed and compelling long opportunity with limited analyst coverage on either the buy-side or sell-side.  (For what it’s worth, two of the three comps we mention below haven’t been discussed on VIC ever, and TE was last discussed in 2012.)  

Because Bel has been so forgettable for so long, there are few eyeballs on it as it is going through a substantial transformation.  Analysis of the aforementioned terrible historical financials is, in this case, less relevant because of the new approach.    

Under new CFO Farouq Tuweiq, who took a pay cut (from being a sell-side analyst covering the industry) to join Bel, the company’s direction is very different than it was before he joined.  This isn’t even a “turnaround” - the company has substantial near-term revenue momentum and good products and technical expertise; the vast majority of its failures over the past decade are due to relatively easily fixable problems with their sales process and compensation structures.

Here’s how Farouq explained it to us in a nutshell: peer electronic components companies both private and public typically operate at circa ~30 - 35% gross margins and ~20%+ EBITDA margins.  If you look at the past decade of margins for these companies, such as Amphenol (APH), TE Connectivity (TEL) and Littelfuse (LFUS) in the public markets and Molex in the private markets, Bel stands out like a sore thumb, with gross and EBITDA margins a minimum of 10-15% lower, and often more, depending on the year.  

(Note that none of these comps are perfect as many of them have divisions that aren’t comparable, such as power semiconductor / sensors products or major automotive exposure, but as a whole they are useful.)

Since 2016, Bel has never had a full-year gross margin exceeding 26%, while its peer CTS - with very similar levels of revenues - has never had a full-year gross margin below 32%.  SG&A efficiency was also worse, since salespeople were being paid for bringing in business that was actually losing the company money.  

As a result, Bel’s EBITDA margins were mid to high single digits at best, while CTS had EBITDA margins consistently in the 15 - 20% range.  

Unsurprisingly, as a result of this poor performance, Bel Fuse trades very cheaply.  Despite actually having higher revenue than CTS since mid-2015, Bel has always traded at a lower valuation - today, with very similar top lines, CTS trades at an enterprise value 3x higher than that of Bel.

We’ll discuss valuation in more detail later, but summarily, this is the investment thesis - as the operating gap between Bel and its peers narrows, so will its valuation discount, leading to a one-two punch of 50-100% higher margins and a 50-100% higher multiple.  In his time on the sell-side, Farouq observed some comparable situations in private equity, and saw Bel as an attractive opportunity.  

His cash compensation is fairly low (base salary under $300K) and the only way Farouq will make substantial money from this position is if Bel’s stock goes up quite a bit - equity compensation isn’t particularly generous in our view (he was provided a 10,000 share grant in 2021, worth $170K at the time, and none so far in 2022). 

That being said, we’re not sure financial incentives are the only or primary motivation here (Farouq seems like he wants to run something.)  In many years of meeting with small-cap management teams, Farouq stands out as unusually thoughtful, competent, and straightforward.  

We think it is difficult to invest in Bel without talking to him; although there is certainly plenty of research to do on the company and industry, we think that this is a really company-specific thesis, and more specifically a Farouq-specific thesis.  We wouldn’t generally be enthusiastic about a sell-side analyst running a company, but in this case we think it is exactly what Bel needs.  

When we initially met with Bel about six months ago, given the history of the company, we initially viewed it as a throwaway meeting that we were taking mostly as a favor to an IR guy who we like.  We were fully prepared for it to be a waste of our time and were skeptical about the story, given the company’s colossally bad history.  

Instead, we were intrigued after doing our initial due diligence, and once we met, we came away impressed enough with Farouq - and his strategy - to take a small starter position, and through a combination of strong returns and further purchases over the past six months, it is now a core position for us.  

While we have tried to lay out the investment case here, we highly recommend speaking to Farouq and judging the merits for yourself.

Brief History + Specifics On The New Strategy

Bel was founded by 26-year-old Elliott Bernstein in 1949 after he came back from the war.  The company eventually went public in 1967.  One of the founder’s sons, Daniel Bernstein, became CEO in 2001 shortly before Elliott’s death, and and remains in that position.  More details on the company’s history are available here. 

His economic interest is modest (a little over 3% of the company) but he has nearly 20% of the voting power along with corporate governance provisions that make it very hard for any outside shareholders to actually exercise voting power.  So, for all intents and purposes, this has historically been run as a family company (and not the good kind, either).

We’ve never spoken to Mr. Bernstein and thus don’t have a strong opinion about him, but we think the results referenced above speak for themselves.  According to Farouq, and supported by our research, there is really no structural reason for the gap between the company and its peers.  

As we’ll discuss later, there are more engineered, higher-spec, higher-margins portions of the industry (such as aerospace, automotive, military, and emerging technology) and lower-spec, lower-margin portions of the industry (such as consumer devices, appliances, and so on.)  

Bel’s products and customers match up with the former, and not the latter - so we believe that Farouq is correct and that their poor performance is not really due to being in the wrong places or selling the wrong products; in fact, they seem to have good products and quite a bit of technical expertise, judging by their exposures and various new wins (such as in the emerging e-mobility space with blue-chip customers).  

We believe the reason for the margin delta is that Bel has simply been extremely poorly managed for a long time, with recent results demonstrating that even a little bit of professionalization is massively improving the quality of the business.  It is of course possible that our analysis is wrong, but based on all the research we have done, the areas Bel operates in, and the products they offer, should be high-margin.  

We want to differentiate the situation from a turnaround (which we tend not to invest in) because the company isn’t broken.  That is to say, they have good products and technical expertise in good markets that are growing; they have a reasonable manufacturing footprint that is not in need of a complete overhaul (although there is some room for improvement), and so on.  They just had no clue about pricing and margin discipline, combined with a questionable M&A strategy.  (If you want to call it a turnaround, that’s reasonable - we won’t push back.) 

At some point, Bernstein grew frustrated with the stock price and realized that continuing with the existing approach was simply not working, so he brought in Farouq to fix things.  When Farouq  joined in February 2021, he was the first external executive the company had ever hired.  Moreover, he is by far the youngest member of the executive team - he was 39 as of the last proxy, compared to all others being in their 50s or 60s.  

While Farouq is not the CEO, he is the one who is driving the strategy and essentially running the company.  To name a few examples, when Farouq joined, he found extremely low-hanging fruit such as the following:

• A lack of any data or understanding of profitability at the SKU or customer level - the company simply had no clue which products or sales were profitable and which ones were not.
  
  

• A company culture that incentivized and rewarded booking sales regardless of margins - for example, the company had a large piece of business with Facebook that was only marginally profitable at best.
  
  

• Long-time clients who were friends with management or salespeople, who Bel was doing business with at negative gross margins (?!) - Bel has subsequently gone to these customers and either repriced the business to acceptable margins, or agreed to walk away from it
  
  

• A lack of operational management expertise
  
  

• A lack of basic sales competence - focus was on acquisitive rather than organic growth and long-term existing relationships rather than signing new business - so terrible gross margins were compounded by poor SG&A efficiency due to unproductive salespeople; it seemed very much to us like an organization that was “coasting” rather than high-performing
  
  

• Doing significant custom engineering work for customers for free - i.e. not charging them higher prices for a more specialized product (again, ?!)

In theory, fixing these problems is not rocket science.  You just change compensation practices and eliminate low or negative-margin pieces of business, backfilling it with business at more acceptable margins, for which the current environment has been favorable, de-risking this portion of the transition over the past six months given the substantial backlog as of today (we were initially concerned that the company might need to reset at a lower revenue level; instead they are reaching all-time highs).  

In practice, of course, people are resistant to change.  So Farouq had to be selective about going after the worst offenders first - i.e. shedding negative and low margin pieces of business and changing incentive / sales comp.

This is why we think Bel is a great opportunity now and we’ve added at higher prices.  Our first and biggest concern was how realistic it was to change and and professionalize the culture at a family company that had been very “sleepy” and non-performance-driven for an extremely long time - i.e. “can you teach an old dog new tricks.”  The culture was one of coasting and we doubt anyone would have even known about, let alone want or support, a Danaher Business System like approach.

When we discussed this six months ago, Farouq acknowledged this to be true and noted that when he joined and even at that time, it was often two steps forward and one step back, with many people in the organization resisting changes.

We think that Bel’s results over the past year suggest that people are indeed getting with the program.  In the first six months of 2022, Bel’s gross margins expanded by 210 basis points and with a similar improvement in SG&A efficiency, its operating margins nearly doubled from 4.4% in the prior year to 8.1% this year. 

Furthermore, we believe that these results actually understate the magnitude of the change, because all the orders Bel is taking today are under their new, profitable pricing regime - but given extended lead times in the industry due to supply chain challenges they still have substantial orders/contracts in backlog from the “before times” when they were willing to accept peanuts.  

So further improvements are, in our view, already “baked in” and we think that Bel’s margins should expand into the low double digits in 2023, even if the company ends up facing potential recessionary headwinds.

In some senses, we think that going from 30 - 60 mph will be easier than going from 0 - 30 mph because momentum is already in Bel’s favor.  

Indeed, Farouq was much cheerier when we talked to him after Q2 results than he was when we met him earlier in the year - he noted that the company has given shares to many employees in their retirement plans, and in addition to the fundamental results, the employees are noticing that the stock is up big in a market where lots of things are melting down.  

(As an aside, Farouq has put his money where his mouth is and continued to buy some shares on the open market as the price has risen, such as a 1,000-share purchase on March 11 at $17.77 - we generally don’t focus too much on insider activity but in this case we think it’s notable.)

We believe that organizational resistance is lessening - most of the people with strong objections have probably left by now, and those who are there are motivated to keep the train rolling since they can now see the benefits of the new strategy in both their retirement accounts and their pocketbooks (now that, e.g., compensation is focused on profits rather than revenues.)

Industry Overview

With that out of the way, let’s discuss the industry.  We think it makes more sense to start here, and then move on to Bel Fuse specifically, because without the industry context, Bel’s product lines and exposures won’t make a lot of sense.

The electrification, networking, and computerization of the world provides long-term secular growth opportunities for electronic components such as fuses, capacitors, connectors, and coils. 

In day to day terms - your car is now a computer on wheels and one Deloitte study suggests electronics are now 40% of the cost.  John Deere now employs more software engineers than mechanical engineers.  Traeger has become a household name in large part because you can press a single button on your phone and the grill will cook a recipe from start to finish, changing and maintaining different temperatures along the way.  

Throw in the transition from internal combustion engines to electric engines, and clearly there are plenty of growth opportunities.  Cavalierly, we’d say that if you take half the buzzwords out there - “5G,” “EVs,” “AI,” “solar,” and so on - all of them will create more demand for boring electronic components that nobody thinks about.  Fuses and connectors that cost a few dollars aren’t sexy and are rarely thought about, but they’re necessary to enable all the technologies of the modern world.

It’s a highly fragmented market with a handful of very large companies (TE, Amphenol, Molex, and Littelfuse are among the larger ones) and a long tail of much smaller ones.  As with any market, there are more and less commoditized parts of it.  It’s a global market in the tens or hundreds of billions of dollars depending how you define it, but not all of that market is a good target for Western companies.  Earlier this decade, TE Connectivity (TEL) walked away from a billion dollars in revenue in highly commoditized areas such as consumer electronics, which they discussed extensively on their conference calls at the time.  

They noted that areas like appliances are low-margin business with generic Chinese competition; it’s all about cost per unit and it’s difficult to differentiate on anything other than price - a battle which a Western firm is not equipped to win at margins and returns on capital that shareholders would consider acceptable.  (Conversely, Littelfuse still has some business in this area, and has strong margins in the segment overall, suggesting there are still attractive pockets even in that vertical.)

On the flip side, there are many applications that are more highly engineered or in harsh environments.  At the extreme, TE Connectivity (which has a big position in automotive) explained the following at a 2021 JPMorgan investment conference:

So what we find is our solutions are very sticky because there's a very -- there's a high degree of certification and qualification that goes into these solutions. And you have to think about them. These are in very harsh environments. They're very robust and the cost of failure of our solutions is extremely high. 

So the cost -- the initial stage of getting engaged. In many cases, we're talking to an OEM 7 or 8 years before where we're actually designing -- the final design is in production and at a high volume.

 And so what ends up happening is that there's a very high level of carryover between vehicles. So you'll have a platform. We get designed in on that platform and then that platform will run 3 to 5 years. 

And then when it comes time for the next platform, there's a good portion of that bill of material that continues to the next platform, and then there's a portion that's innovated. And what we find is that the value of incumbency is extremely high. 

And that's why initial incumbency is really important because just like we have a large moat around the positions that we already have, competitively, it's very difficult to access that business once -- going in with a 3% or 4% or 5% lower price doesn't change the sourcing decision of an OEM because the weight of replacing that component is -- it's just too high. The risk is too high relative to the return.

A former employee at Amphenol made similar comments on another expert call, suggesting that parts of the interconnect space are a lot stickier and less transactional than an outsider might assume, contributing to the revenue and margin stability:

And of course, TE Connect has certain customers where Amphenol doesn't get into, or Molex. They all have the stuff, but that's where I'm thinking is very strong loyalty.

And the other piece, once you enter these accounts, you have almost the manufacturing environment behind it, because even though many of the parts you produce are considered more or less the standard, which is true, it could be done by others as well. But once you really have the infrastructure in place to actually make many of these parts, it's pretty cost prohibitive for both customer to switch or for a new supplier to come in and take it over.

So that's why it's kind of an interesting scenario. It's like once you're in, then you have probably pretty much a long-term kind of really an annuity type kind of relationship with these customers. So long answer, very loyal, and they make really hard to maintain it.

We don’t think that this is necessarily applicable to every single application for every single industry, but electronic components do seem to fit into the “low cost, high cost of failure” model that we like for industrials.  

These components are generally not very expensive relative to the cost of the end-product, but are absolutely critical to its function.  The vendors (including Bel) have significant engineering staff that works with customers to spec their product into customer designs, providing some level of stickiness/moat in the near term.

For example - and again, to be clear, this is one extreme, Bel acquired RMS, which pre-COVID was doing about ~$45 million of annual revenue supplying into commercial aerospace (with strong exposure to narrowbodies such as the 737 MAX and A320neo).  

Comparatively, the price of the smallest models of a brand new 737 MAX or A320neo is generally about the same ($45 - $50 million), and Boeing and Airbus each did $70 - $80 billion of revenue in 2019.  RMS’s entire revenue selling into all of the 1,500+ commercial airplanes produced in 2019 was the same as the cost of a single 737 or A320.  

So clearly one can mathematically determine that the cost of RMS’s electronics are a completely insignificant rounding error to the cost of producing a commercial airplane.  Even a savings of 20-30% would not impact the bill of materials in any material way.

But looking at where they are used (see the Bel Fuse slide below), clearly you can’t afford a failure in areas that their products go into such as “wheel hydraulics, flight controls, navigation, mechanical systems, fuel quality, and navigation controls.”  

Even setting aside the regulated nature of aircraft production, particularly after the high-profile - and costly - safety failures with the 737 MAX, I doubt Boeing and Airbus would risk swapping out a component in the flight control and wheel hydraulics systems for the sake of saving a few dollars on a new aircraft.  

An expert call transcript we reviewed on Littelfuse (LFUS) made similar points about how in areas such as electric vehicles, DC charging, and solar/wind farms, higher-spec components are needed.  This requires R&D / engineering talent, providing opportunities for Western companies to differentiate from their generic Chinese competition and earn an acceptable margin.  

The expert:  

With the electric vehicles, they have to have a certain level of vibration testing and everything that regular electrical fuses don't have. [...] 

When you think of electric vehicle, there's significantly more electrical components that need protection and, like you mentioned, a much higher amperage and DC voltage and everything. They're converting that electrical energy into motion, so requiring those electric vehicle charging and DC charging fuses, which are definitely more robust. 

I believe their components are made with not the glass little pieces that you think of when you think of automotive sensors and fuses. They're more of the, what's the material, the copper contact and the center part is a more robust material than glass.

[...]

There's a lot more to that that I know is being developed and, with the motion of vehicle, those fuses are a lot more sensitive. Typically, they have a higher level of overcurrent sensitivity, which requires more intricate level of testing. I didn't dig too deep into it, but I did work a little bit with the product managers to understand that that was an area that they still were developing. Also, DC charging stations and electrical storage for solar and wind farms.

Again, these are just examples, but the point is that there is plenty of market opportunity in the higher-spec areas.  (We will discuss Bel’s product portfolio momentarily.)

Obviously, we believe that some of what is today high-spec, leading-edge technology will eventually be commoditized by Chinese competitors over time, but Western firms with engineering expertise will continue to evolve their products and stay on the leading-edge, more technologically-demanding applications.  

While Littelfuse (LFUS), Amphenol (APH), and TE Connectivity (TEL) admittedly have other business areas such as power semiconductor and sensors - and are therefore not perfect comps - we would point out that they have $30B+ of combined trailing revenue at a roughly 23% EBITDA margin.  CTS is also the size of Bel Fuse (roughly) with, as we discussed, a consistent history of 15-20%+ EBITDA margins - demonstrating you don’t have to be at multi-billion-dollar scale to generate those kinds of margins.  

Thus, it’s difficult for us to believe that in an electronic components market that is in the hundreds of billions of dollars (depending how you measure it), Bel Fuse won’t be able to carve out $750 million in revenue at attractive margins, which is all we need for the stock to double or triple.  Indeed, there are already big chunks of their business that operate at good margins - their historical struggles were caused by lack of discipline and management competence rather than lack of opportunity.

We would finally, somewhat briefly, touch on cyclicality - there can be cycles both with specific end-customers (military and data center spend can go up and down depending on product releases, for example) and there can also be broader electronics industry cycles - Littelfuse, TE, Bel, and everyone else saw a softer 2019 due to destocking in the distribution channel, among other factors; Littelfuse saw a decline of 13%, TE saw a decline of 4%, and CTS saw a decline of 2.3% (demonstrating that specific industry/customer exposure is important).  

Littelfuse did note that there are typically five-year cycles in electronics and the past ten years prior to 2019 were unusual in that the upcycle was so long.  It is possible that the increasing electrification of the world today, including more electric sources of power and more “connected” devices, will cause historical cycles tied to e.g. consumer product releases to be somewhat less meaningful than they are today.  

As an example, Littelfuse highlighted on their call this month how they have grown their passenger vehicle business at a high single digit CAGR over a three-year period with a high-single-digit annual decline in global vehicle production, due to substantially increased content opportunities:

Since 2019, our passenger vehicle business has grown high single digits on a compounded annual basis, while global car production has declined high single digits. We see a number of ongoing content growth opportunities with electrification, electronification and ADAS and expect to continue our long-term market outperformance

Bel doesn’t play much in auto, but we think the idea holds - as a hypothetical, even if (for example) John Deere produces 20% fewer tractors in two years than it does today, they could easily need 50% more electronic components per unit for the increasingly electrified tractors that they do produce, resulting in Bel’s revenue being 20% higher.  That’s not to say year-over-year revenue declines will never materialize, but over a multi-year holding period, secular trends should likely outweigh cyclical fluctuations.

Ultimately, we don’t think cyclicality is extremely severe - in fact, it seems fairly modest relative to, say, certain other types of industrials, or semiconductors.  At the moment, some of Bel’s customers are actually coming out of the COVID-driven downcycle, as we’ll discuss in a few sections.

Bel Fuse - Business Overview

Within the context provided above, let’s discuss Bel Fuse’s business segments specifically.  There are two ways to categorize it - by product, and by end-market.

From a high-level, Bel’s sales are broken down as follows:

• 14% military and aerospace

• 18% industrial and EV

• 36% network and cloud

• 32% “distribution” (a broad range of end markets where they don’t have visibility to the end customer - an example can be found here on Avnet’s website)

We’ll look at the individual segments briefly.  The company’s “Magnetic Solutions” segment (30% of revenue) is where the majority of their network and cloud revenues are, including a strong relationship with Cisco.  

This is a somewhat lower gross margin business (23.9% LTM) and will always be lower than the other segments (although there is room to improve), as many customers are large networking/telecom providers (77% of revenue in this segment is from network) who do significant volume, but SG&A is also lower here due to the more concentrated nature of customer relationships.  They have a strong position here in “integrated connector modules” that I would highlight as it demonstrates their innovation / engineering competency - it combines magnetics with an Ethernet connector, thereby lowering the footprint needed (allowing customers to save valuable space in servers / data centers). 

This is a product line where if you Google the term, Bel dominates the first page and a competing product from Pulse Electronics doesn’t even appear to show up until the 14th or 15th result.  Growth in 5G and cloud spending as a whole should benefit the company, although as I referenced previously, the company has walked away from high-volume / low-margin business (such as Facebook) and is focused on more niche applications with higher margins.  (I would note that Amphenol has business with Google, Facebook, and Apple at apparently good margins, per one expert call, but they are also much larger and may have scale advantages that Bel does not have.)  

One risk factor to note here, as the big 3 cloud providers increasingly take share, is that Bel is levered to the lower-growth portion of the market - so we wouldn’t expect Bel’s revenues here to approximate the hypergrowth of, e.g., Azure or AWS.  But our thesis isn’t really based on revenue growth and in fact, we think Bel will work out just fine even if revenues are flat or only grow modestly.

Anyway, he company’s “Power Solutions and Protection” (40% of revenue) - which includes the “fuse” part of Bel Fuse, as well as power supplies and converters/inverters - also has some networking exposure (29% of revenue), but also services a broad range of industrial customers .  A particularly exciting exposure in this segment is the rapidly growing “e-mobility” piece, which is run-rating at $7 million per quarter but growing rapidly (it has nearly doubled year on year, and has substantial backlog - bookings increased by $47 million in 2021, for some context).  

While the company does have some business with Tesla (TSLA), it’s not really focused on the automotive space or “standard” electric vehicles, because they feel like that’s not really a customer set they are equipped to win.  Rather, they are looking for nichier, higher-spec opportunities (this holds true across the rest of their business as well.)  This includes things like electric tractors that John Deere is developing, electric Post Office vehicles, specialized mining and other off-road trucks, and so on.  

They have a substantial incumbent position from having been in this business for 10+ years, and are targeting a 30% market share of the applications they target, which they estimate will be a $300 - $400 million TAM in three years, which would translate into fairly sizeable revenue ($100 million) relative to our estimate of their forward revenue ($650 million annually). 

To be clear, growth in this segment is not required for the thesis to work, but it does seem likely (see the aforementioned expert call on Littelfuse) and is a nice kicker. 

The final segment is “Cinch Connectivity Solutions” (another 30% of revenue).  This segment has high performance (including harsh environment) connectors, fiber optics, and so on.  For those interested, the August 2020 Amphenol writeup by Jumbos02 contains some background on the connectors market and some context on Amphenol which is broadly similar to what we’ve discussed above for the electronic components industry writ large.  

Connectors do what it says on the tin - they help connect something to something else.  Think of the Ethernet or USB port on a computer - those are visible connectors (which Bel sells), but there are of course plenty of invisible ones in the guts of electronics as well. 

In terms of the value proposition of interconnect systems, Amphenol noted the following in a conference call a decade ago, which we think is only more applicable today given the electrification trends previously noted: 

 power requirements of IT systems, of cars, of military hardware, whatever that may be, industrial hardware, are increasing and are creating needs for more efficient interconnect. Turns out that the interconnect in a product can be one of the core areas of loss of the power.

As it relates to Bel specifically, about 40% of revenue in connectivity is from distribution, and other than that, their biggest end-markets here are military and aero, which are very high-spec, and thus high-margin, from what we understand (the arguments we made earlier about commercial airplanes apply equally well to weapons or defense systems, military aircraft, or so on).  

Aero represents a source of upside on both revenue and margin as commercial plane volumes recover - pre-COVID volumes were 50% higher ($45 million) vs. current run rate (low $30s) based on the most recent quarter, and margins are also currently depressed due to timing of hiring and ramp-up.  However, the company is confident that this will improve into next year.

This segment is the only one where LTM gross margins are lower over the LTM than the previous year, a large portion of which (in the most recent quarter) is attributable to inefficiencies related to rapid ramp-up with commercial aero customers (such as training costs, etc) that should stabilize going forward (this is structurally a very high margin business).  

Farouq noted that some of the same steps they’ve taken in other businesses are being taken in Cinch, and they expect significant margin improvement over the coming years.  Some of this is related to the timing of customer contract renewals / renegotiations:

Power and magnetics will see better year-over-year improvement, while connectivity will be slow to follow until year-end as contracts come up for negotiations

Ultimately, we think that Bel’s business is reasonably well diversified and mostly exposed to positive longer-term trends - defense spending can be cyclical depending on programs and we would highlight that their networking exposure is not likely to be as high-growth as Azure/AWS, but we don’t see any dying or structurally commoditized parts of the business.  

It seems like a solid base of revenue from which to grow upon; their mix might look very different in several years thanks to emerging technologies (e.g., their e-mobility piece seems likely to grow to 15% and perhaps up to 20% of revenue in a 3-5 year time frame).

Current Business Trends & Outlook

Business is booming right now.  For the three and six months ended 6/30, revenues were up 23% and EBITDA was up 64%/100%.  The margin improvement in the business is noticeable but according to Farouq, it’s still “early days” with lots more to come.  On the earnings call, he put it this way - 

We think that we have a healthy room to go yet on the margin front. And we don't obviously put on any forward guidance, but we understand that we're still kind of climbing the stairs here early on. 

In a follow-up call, when we asked what provides them with confidence that they have room yet to go, Farouq noted to us that there’s still plenty of room on the same initiatives - 

Where it’s coming from is being more strategic and purposeful with the business we’re going after - negative gross margin, raise or walk away.  Also, business we are booking is coming in at a higher margin [author’s note - he means relative to what they are reporting and what is already in backlog from previous quarters] - we’re using our R&D to go after better business and get better markups on the front end.  We’ve also taken share from competitors, and have a big focus on doing better - good engineering, have to be more aggressive pounding the streets.

In terms of the forward outlook, backlog is $580 million, up 24% from the end of the year in line with revenues.  Even allowing for some possibility of double-ordering by customers and a higher book-to-bill than usual given longer lead times, the company sees no signs of softening demand despite the well-publicized macroeconomic concerns. 

Indeed, they had another $34 million in orders that were supposed to ship in the quarter (relative to $170 million of revenue) that they were unable to due to supply chain challenges; as such, demand vastly exceeds their ability to fulfill it, softening the blow of any future slowdown.

It’s difficult to compare results apples-to-apples with peers due to different exposures, but Bel’s near-term results are at the high end of the peer group, suggesting they are doing something right.  Part of the delta appears to be that Bel is not exposed to two categories that are slowing down - personal electronics / consumer devices (demand for which was pulled forward during COVID) and auto production (which has been hampered by the well-telegraphed semiconductor shortage).  

Bel might also simply have an easier time since it is smaller than most of its peers and thus winning $10 - $20 million of new business still moves the needle for them.  Nonetheless, the reinvigoration of the sales force appears to be paying off with this strong organic revenue growth.  

While they acknowledge that a slowdown might affect parts of their business in 2023 (such as data centers, where Amphenol is already seeing slight softening, and we already see some indications of slowing from other companies we follow), we think they stand to benefit from cyclical recoveries in some areas (like aerospace) that are still depressed post-COVID, and secular growth in other areas (like e-mobility) which we discussed previously.

The focus continues to be on winning new business in attractive areas at attractive margins; while the company does look at M&A, Farouq is both operationally realistic and financially savvy.  

He has told us on multiple occasions that they might do a small tuck-in here or there if they can get it at a really good price, but the organization is not yet ready to take on significant integration (get your own house in order before you buy someone else’s).  

There are quite a few opportunities, though; as we mentioned, the industry is highly fragmented, and consolidation logic makes sense, as you’re able to gain a higher share of wallet with the same customers, creating synergies and increasing stickiness - all of the big players (TE, Amphenol, Littelfuse) have grown significantly through M&A (see Jumbos02’s APH writeup for a good example in the connector vertical - the logic applies to electronic components more broadly).  Bel, like CTS, is kind of in the middle; it’s not too large for someone else to acquire it, but it’s also large enough that it could likely add significant pieces.

As you would expect given Farouq’s background, he’d rather wait to do M&A until he can exploit an equity arbitrage (i.e., with his own stock trading at 5-6x EBITDA, he sees no reason to spend capital on things at higher multiples - the math gets easier if and when Bel rerates to peer multiples.)  

So, we think that the company will continue to shepherd cash for now and capital allocation isn’t a hugely meaningful near-term part of the story, but something to return to in 2-3 years once the operational issues are fixed.

Valuation

Relative to our typical holding, there is a wider range of potential outcomes for Bel because their business will obviously look very different in 2-3 years if Farouq is successful - the question is how different.  The comp set has a fairly predictable and stable margin range but it is obviously difficult to know whether Bel will end up at the low or high end of this range, given that there is not really any company-specific financial history that is relevant.

But even under conservative assumptions, there is tremendous potential upside here.  So we’re simply going to present what we think is a realistic upside case, and then leave additional potential to be discussed later if and when the company executes on the first part of the story.

Starting with the basics, the company has 10.4 million class B shares (non-voting) and 2.1 million Class A shares (voting) for about 17.5 million total (rounding).  Class B shares currently trade at a slight discount, although they have historically traded at a premium.  For reasons we’ll discuss below, we don’t think voting rights can actually be used here, so we would prefer whichever class is cheaper (with liquidity generally guiding us to Class B unless the A discount was very substantial.)

For analytical purposes, the shares are more or less economically equivalent - our reading of the 10-K (see exhibit 4.1) states that B and A shares get the same consideration in a liquidation or acquisition of Bel, but if cash dividends are declared, B shares are entitled to a 5% higher cash dividend than A shares.  Currently the dividend is $0.07 quarterly for class B and $0.06 quarterly for class A, but nobody is buying this for the dividend and we don’t think the difference is that material.

Anyway, based on 12.5 million diluted shares out in total and a ~$25 price for the B shares, the company’s market cap is a little over $310 million.  (We do know that the technically correct way would be to add up the value of the A shares and B shares, but since the B shares are actually fundamentally worth more than the A shares, and have historically traded at a premium, we view the current premium of the A shares as a market-inefficiency artifact and just value the company based on the B shares.  The difference would only be a few percentage points anyway.)

The company has $112 million of long-term debt and $65 million of cash for net debt of about $50 million or less than 1x LTM EBITDA.  We would note that given extended lead times and the supply chain challenges everyone knows about, inventories are currently substantially elevated, and inventory turns are lower than historical levels.  

In a more normalized environment, we believe this would be an additional source of cash, although not one we are including in our model.  As we discuss below, in the longer term, we also believe there is room for Bel to optimize working capital relative to its historical levels (approaching peer metrics), which would result in a further release of cash.  Anyway, for now,the total enterprise value is roughly $375 million at the current time.

On an LTM basis, the company has generated $600 million in revenue and $59 million in EBITDA.  So to start off with, the current enterprise value is only about 6 - 7x LTM EBITDA.  That’s not all that expensive.

But it gets better.  Due to the rapid trajectory of margin improvement, as well as strong revenue growth in many areas rebounding from COVID, in the most recent quarter, EBITDA was $19 million on a revenue base of $170 million.  

There is some seasonality to the business due to the Chinese New Year and holidays in North America (Q1 is typically the weakest quarter and Q4 is typically a step down from Q3).  So annualizing Q2 numbers is likely overly aggressive.  Conversely, as we discussed, gross margin improvements for the next year are already baked in.  

So we think that forward revenue and EBITDA are closer to a figure of $650 million and $70 relative to the LTM figures of $600 and $58, which would put Bel’s valuation at about 5x forward EBITDA.  (Note that the ongoing margin improvement initiatives also represent a source of revenue growth; if Bel earned a 20% gross margin on a $10 widget and raises prices to achieve a 30% gross margin, revenue must grow by 10% at flat COGS to accomplish that.)

We think that the eventual outcome here is that Bel manages to raise its EBITDA margins to the 15%+ range, more in line with peers - the company has already demonstrated solid progress and Farouq is only 18 months in, and has now earned respect within the organization to go after the meat of the problem.  

Let’s say that in three years, revenues are flat but EBITDA margins have reached the 15% level (the underlying assumption being that Bel is currently less focused on absolute top line growth but rather walking away from or repricing lower-margin business and replacing it with more attractive high-margin business, and also that the macro environment might be more challenging).  

Bel would then be generating over $100 million in EBITDA, and we think that once it is a “regular-way” company (as Farouq puts it), it will trade at peer multiples.  The aforementioned comps have generally traded at a minimum of 8-12x EBITDA over the past decade, and we believe Bel can achieve at least 8x - this is not a hugely capital-intensive business, although we’ll address ROICs in a moment.  

We exclude Amphenol from this comp set analysis because it has a unique, somewhat Danaher-sounding culture and highly diversified business model that limits its cyclicality, along with an active and successful M&A program; as such, it is viewed as a true “compounder” (again, see the analysis by Jumbos02) - it has traded at over 15x trailing EBITDA most of the time since 2014.  For an additional data point, Koch Industries acquired Molex in 2013 for 11.2x EBITDA.   

The relative valuation is just a sanity check and we almost always prefer absolute valuations.  At Bel’s capex and working capital intensity, fully taxed FCF at a modest growth rate would probably be in the low $50 million range, so ~8x EBITDA translates to a ~14x free cash flow multiple at most - not very aggressive for a company with long-term secular growth opportunities and continued margin expansion opportunities (since it would still, at that point, be at the low end of the peer group).  

Free cash flow conversion would further improve over time as additional improvements were made to capital efficiency (which as we stated, is not the near term priority), justifying a higher EBITDA multiple and thus enterprise value at the same free cash flow multiple - we think that Bel will not deserve as high a multiple as its peers until it has comparable operating performance. 

All that said, 8x EBITDA on 100 million in EBITDA would value Bel at $800 million or $64/share compared to $25 today, compared to the current EV of $350 million, suggesting about 150% upside for the stock (interim cash flow would pay off current debt, so the $800 million would be entirely market cap.)  But even this is conservative; CTS is valued at $1.3 billion today with slightly lower revenues than Bel, so if Farouq can close the margin and valuation gap to CTS, then there is even more upside than we are suggesting.

One note - we acknowledged that Bel has historically had terrible ROIC.  This is not really a function of the business being inherently capital intensive (capex to revenue averages only 2% for Bel); peers achieve acceptable double-digit ROICs in line with other industrials.  

Bel made some terrible acquisitions, but an even bigger problem is their lack of profitability.  Higher profit margins fix the numerator part of the ROIC equation, both by increasing profits, and by reducing the amount of inventory required to service the top line (adding insult to injury, Bel had capital tied up holding inventory for these low or negative-margin customers).  

As for the rest (i.e. the denominator), Farouq acknowledges that Bel’s working capital efficiency (inventory turnover, etc) is below that of peers - however, this is more a further-out component of the strategy.  

Their ROICs will look a lot better once the margin structure is fixed, and operational improvements at the factory level are the next stage of the process.  There is of course some heavy lifting to be done here, but we think Bel is so deeply discounted that the risks are more than priced in - Bel doesn’t have to become an operational rockstar to double or triple in value over the next three years.  

If the margin improvement strategy works, then ROICs will be an acceptable (although not spectacular) low double digit level, with a pathway to becoming the more teens plus type levels that peers (and good industrials generally) achieve.

Risk Factors

While not exhaustive, there are three risk factors we see fit to discuss.

First is the widely telegraphed potential for a near-term economic slowdown.  Bel is not currently seeing this in its numbers (in fact, it has record backlogs).  However, Farouq acknowledges that it is certainly possible that a slowdown could manifest in 2023 (there are some indications, for example, that IT spending is already slowing, which could affect their Cisco / Magnetics business).  We think that mitigating factors here include the following:

1. Most importantly, the ongoing margin improvement initiatives - many of which, again, are already “baked in” (new orders have higher margins than existing backlog) - will cushion current profitability against economic headwinds.  There is a realistic scenario in which the macro is bad, Bel’s revenues shrink modestly, but Bel’s EBITDA continues to climb as EBITDA margins go from single digits to double digits to teens.
   
   

2. There are significant pieces of the business that either have strong secular tailwinds or cyclical recovery stories.  For example, the company’s aerospace business is ramping back up after the dual shocks of the 737 MAX grounding and COVID, and to a lesser extent other issues such as supply chain challenges for the A320neo, production halts for the 787, and so on.  As such, even in a strongly recessionary environment, aerospace revenues and EBITDA should increase meaningfully in 2023 relative to 2022.  Similarly, the company is seeing rapid growth in its non-automotive electric-vehicle applications as previously discussed, and this secular trend should continue to drive growth even if total vehicle production volumes decline.
   
   

3. Sales have been deferred and costs have been increased by ongoing supply chain challenges; Bel has been unable to fulfill all demand that it currently has from customers.

The second major risk we see is the ownership structure.  Keep in mind that while we believe Farouq is functionally running the company, Bernstein is still actually in charge, both as CEO and due to his voting power.  There is a two-class share structure (we own the non-voting BELFB because it is cheaper and economically equivalent to very modestly superior to the voting-class BELFA).  

Bernstein controls over 20% of the voting “A” shares although only owning a little over 3% of the total economic interest in the company, and there is a mechanism in place that prevents outside parties from exercising voting rights unless they pony up to buy an equivalent stake in the B shares.  Page 15 of the 10-K and page 5 of the last proxy describe how for this reason, GAMCO (Gabelli) had no voting rights despite owning 19% of the voting shares as of that time.

So, Bernstein clearly hired Farouq to change the company and Farouq appears to have his full support.  Recent stock performance is certainly an incentive for Bernstein to keep following this path.  But change is hard, and we can’t rule out the possibility that Bernstein changes his mind at some point, for some reason (stranger things have certainly happened) - and in that case, shareholders have no recourse.  Similarly, the company’s capital allocation history under Bernstein has been abysmal and he could decide to overrule Farouq and do something stupid.

Third and finally - probably the biggest risk factor in our view is the company’s exposure to China.  The majority of the company’s manufacturing footprint is located in China, and as such any disruption in China would be bad.  In the near term, substantial shutdowns related to China’s Zero COVID policy would disrupt the supply chain, and potentially their operations.  In the longer term, if geopolitical tensions were to lead to tariffs or significantly strained business relations, that would obviously hurt Bel.  

The company does have some manufacturing elsewhere (Slovakia, Mexico) and has looked at alternatives in Southeast Asia (Vietnam, Malaysia, the Philippines) - but diversification is fundamentally difficult because in many instances, their suppliers, customers, or both are located in PRC.  So locating factories elsewhere would increase logistics costs while not really insulating them from the risk factor because of the exposure of their supply chain and customers (i.e., if relations deteriorated completely, hypothetical factories in Vietnam would likely still not be able to receive or ship parts to China, so it wouldn’t really help.)

One can perhaps take some comfort in knowing that this is a) an industry-wide issue rather than exclusively a Bel-specific one, and b) a risk to many, many companies (for example, Apple is 90% reliant on China and would be completely screwed in this scenario).  Nonetheless, it is something to be aware of.

On the whole, while we are not dismissive of these risk factors, we don’t think that any of them are inordinately concerning (relative to other prospective investments, at least) and are comfortable with all of them in context of position sizing.

We look forward to the VIC community's feedback!

Catalyst

The self-help story should continue to drive earnings higher, resulting in Bel being viewed as more of an investable company.